I think I'm following some people and their alts, because they have similar mannerisms and common interests, but then I wonder how much of that is just crypto bro echo chamber.
I'll be cruel to their creators instead.
@AnthropicAI, as a paying user who sometimes gets good results from your systems and often doesn't, you need to grow up and realize how often your non-sentient, non-aware bag of matrix math ruins your customers' day. Humility, now!
splitXCH retweeted
That's a bold assumption, made under our current maths models, and certainly hopium,
It's like saying if aliens come to Earth don't worry they are friendly,
Should we take the alien risk seriously? No
But if they come, will they be friendly? No
Should we prepare bunkers for aliens? No
Same thing
This latest fear of LLMs hacking hash algorithms is fascinating to me. I was part of a collective doing distributed computing challenges to find hash collisions by brute force in the late 1990s. Any valid math attacks to reduce compute space significantly would be destructive.
Hard to match the thrill of discovering a working exploit chain without AI. Really glad I got to do that a few times before the LLM takeover. Ethically, of course. Although it was most fun before I reached the age of majority and faced essentially no consequences for research.
It's really fun to dunk on people for Dunning-Kruger things. Then, sobering to recall that statistically there exists something we're 100% convinced we understand are instead absolutely wrong about and have no capacity to learn correctly.
Anyone know the derivation path for the BLS recovery key for Chia Cloud Wallet? Yes, I know I can just start running values through wallets and comparing them, but this seems like it should be documented somewhere for future-proofing.
It's interesting to me when I read about faster encryption in any context as a benefit, because it seems highly correlated with the encryption algorithm going into decline. I feel bad for people who invented crypto in the 90's trying to build for decades they couldn't imagine!
“7.5x faster encryption in Go with the new "simd" and "simd/archsimd" packages” by Sylvain Kerkour—kerkour.com/golang-chacha20-…
Updates on the 'rust' branch.
Mostly just tweaks to recover from stale coin data. Probably could be cleaned up further.
github.com/trgarrett/xkv8/tr…
A great thing to point your local or cloud LLM at yourself, but I will share the updated branch when I've had a chance to review it. Very likely it was from a workaround I'd added when coinset.org was consistently giving me stale data on unspent coins that were spent.
